Women with newly diagnosed breast cancer who get an MRI scan wait about three weeks longer before their surgery and are far more likely to get a mastectomy than women who have only a mammogram, U.S. researchers said on Saturday.

"MRI may not be as good as we think it is," said Dr. Richard Bleicher of the Fox Chase Cancer Center in Philadelphia, who presented his findings at the American Society of Clinical Oncology's Breast Cancer Symposium in Washington.

"Those who received an MRI had a three-week delay in the start of their treatment," Bleicher said in a statement.

"In addition to the treatment delay, we're concerned that the well-documented false-positive rate with MRIs may be leading -- or misleading -- women into choosing mastectomies."

Bleicher said many women with newly diagnosed breast cancer, including younger women, are getting MRI exams in addition their mammograms. He and colleagues wanted to see if the tests had any impact on a woman's care.

They reviewed the records of 577 breast cancer patients who had been evaluated by a radiologist, pathologist and a surgical, radiation and medical oncologist. Of these patients, 130 had MRIs before surgery to remove their tumors, and 27.7 percent of these had a mastectomy. In the non-MRI group, 19.5 percent had a mastectomy.

After adjusting for tumor size, they said women who had gotten an MRI were 80 percent more likely to get a mastectomy.

FALSE POSITIVES

Bleicher said it is not clear why these women had mastectomies rather than lumpectomies, but it may be related to the higher sensitivity of the MRIs, which are known to have a high number of false-positive findings.

"Rather than having a biopsy to see if those findings are real, women and their doctors may choose mastectomy out of an abundance of caution," Bleicher said.

These findings reinforce a study earlier this year that also found MRIs increase the chances that a woman would have her breast removed.

The research team looked at the time from diagnosis to cancer treatment and found the women who got MRIs waited 22.4 days longer to have their surgeries.

"I can't tell you that a three-week delay will influence her survival, but I can tell you a three-week delay will increase the anxiety on her part," Bleicher said in a telephone interview.

And they looked to see if the MRI exams did a better job of helping doctors predict which women could have breast-sparing lumpectomies, in which only the tumor is removed. They found MRI was no better than conventional mammograms at this.

"We were very, very surprised because people tout MRI scans as being able to see the tumor better. Despite that, we are not able to choose the patients any better with it," he said.

Bleicher said so far there is little evidence to show the scans, which cost about 10 times more than mammograms, help save lives.

"MRIs are valuable and should be done in certain women at high risk, but they are not appropriate in routine evaluation of breast cancer," he said.

Breast cancer is diagnosed in 1.2 million men and women globally every year and kills 500,000.